Fake reviews detection in e-commerce using machine learning techniques: A comparative survey

Abstract

In the field of online commerce, customer reviews have great importance because they significantly influence the profits of a business. Most consumers on the internet rely on reviews to help them make decisions about what to buy because they provide a reliable way to read other people's opinions about a specific product. Since a company's reputation and profitability are directly impacted by the reliability of its online reviews, some business owners pay spammers to create fake reviews. The creation of fake reviews that influence consumers' purchase decisions is a persistent and detrimental problem. Therefore, developing techniques to help companies and customers to distinguish between genuine and fraudulent reviews are still an important but challenging task. As a result, this paper provides a survey on various machine learning techniques are proposed to deal with the problem of detecting fake reviews, as well as the performance of different techniques in spam review classification, and determine the features, strengths, and weaknesses of those methods that may require more development.